INDEX

Explanations

specific terminology related to clinical practices and guidelines

oai_token-act-pair · gpt-4o-mini Triggered by @bot

planning what to test

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

GenerationType

-0.56

complexContent

-0.54

ParallelGroup

-0.42

parsedMessage

-0.40

Espèce

-0.37

gatsby

-0.36

useNewUrlParser

-0.36

 LoginComponent

-0.35

 опро

-0.35

 noten

-0.35

POSITIVE LOGITS

 estekak

0.57

getMenuInflater

0.44

 propagate

0.44

 prega

0.41

Behold

0.41

sizeCache

0.41

 EClass

0.41

 nakalista

0.40

webElementGuid

0.40

 trickle

0.40

Activations Density 0.078%

specific terminology related to clinical practices and guidelines

planning what to test

No Comments

No Known Activations